facebook

The QTLeap project in support of multilingual customer service with quality machine translation achieved its objectives for the first year

logo projectQTLeap - quality translation by thorough linguistic and engineering methods is a joint project funded by the European Commission, between eight partners from Bulgaria, the Czech Republic, Germany, Portugal, Spain and the Netherlands. From the Bulgarian side the partner is the Institute of Information and Communication Technologies at BAS (http://www.iict.bas.bg/). The scientific advisor is Assoc. Prof. Dr. Kiril Simov.

The project started on 1 November 2013 and will last 36 months. Its main objective is to explore and produce sound and innovative methodology for machine translation, which deals with new solutions and using in-depth linguistic and engineering methods for achieving quality translations. The project reached its goal for the first year - a system of automatic translation in a situation of actual use, namely the IIT service in support of HF, Higher Functions. The Portuguese company Intelligent Information Systems Ltd is also one of the partners of the consortium.

Through this service, if a user of an information device or service has a particular problem, he can ask a question in a chat channel for help. If the database already has a similar question, the response is immediately passed to the user. This process helps to minimize human intervention that becomes necessary only in cases where the database doesn’t have appropriate question-answer pairs.

The application of the machine translation system further develops this auxiliary service, which allows you to set the user question in multiple languages. Users ask questions in their own language, and they in turn, are translated into the language of the stored database of questions and answers. The stored response also translates to the language of the original question, so that the help desk can answer in the user's language.

The assessment of the expanded by machine translation functionality Help Center shows that even the base machine translation system achieves a remarkable reduction of the need for human intervention by an average of 60% for each new language that the system supports. This technical progress presents a number of benefits for such companies in connection with the expansion of the general digital market, as well as their improvement in view of productivity and resource optimization, which in turn reduces costs.

The QTLeap project is funded under the Seventh Framework Programme of the European Community for research, technological development and demonstration with financing agreement number 610516.

For more information and contacts: http://qtleap.eu/